I was taught that sexuality follows a gaussian distribution with directional, disruptive, or stabilizing selection. You can find the articles on JSTOR or maybe EBSCO.
WARNING! Wall-o-text!
The above is typical of distributions, ignore the numbers. The integral (the shaded areas) represent the population... thus shading everything below the distribution curve.
The horizontal axis is suppose the represent degree of orientation. Suppose that the far right (-3σ) represents those who are purely homosexual, and the far left are those who are purely heterosexual. Those that are in the middle can be described as bisexual. Basically your orientation can be plotted in that graph somewhere. Perhaps you are in the mauvey-lavender gay area or maybe you are in the bluish center. But this is just the start of the theory, it gets a lot more complicated... as your point, along with the graph's shift.
There are two assumptions to this theory, controversial to say the least.
1) Humans are all born on a moderately neutral "distribution graph..." meaning, a baby in China is just as likely to be in the bluish area as a baby in Scotland... but with some regional genetic differences. (lets just assume that all are the same... okay?)
2) The graph, as well as your position on the graph, change dynamically according to life experience/culture. This means your "perceived" orientation can change. But the "struggle" to maintain that orientation is the difference between the positions.
There are three ways the graph can move...
The above image uses mice as an example... this does not correlate well with sexuality since simple genes are responsible for color... sexuality is more hormonal... but regardless, it will serve this purpose. Also... a mice's color cannot change, but a person's perceived sexuality can.
Stabilizing selection can be found when mice, for example, look at the image. The light-tan ones can find themselves in an environment of dry brush that is identical to their color. Thus, really bright mice in the brush stand out very well... along with dark mice... and get eaten. The tannish mice benefit since they can camouflage in the dry brush, and are more likely to reproduce. Culturally, in regards to sexuality, this could mean a push for bisexuality. A good example of this would be the aristocratic Ancient Greeks (the poorer Greeks didn't really perform too much homosexuality). While the rich Ancient Greeks may have indulged and even encouraged in homosexual activities (pederasty for example, between a mentor and a pupil), they still pushed for heterosexual unions.
Directional selection occurs when one extreme is favored over the other. In regards to the mice, a factory in the area could spew ash over their habitat and make light mice stand out... and get eaten, while there is more chance for the darker mice to live (camouflage). A great example of this in regards to sexuality is... Michele Bachmann.
Sometimes, disruptional selection can occur. What if the ground the mice live on is patchy with light and dark colors? The extremes benefit but the median does not. This is kind of true in reality, as many bisexual men/women/giraffes are forced to pick and choose... (as much as I like mmf or mff threesomes... they don't work).
There are two graphs that are important, the human "birth graph" and the individuals "culture" graph. ''
There were some big criticisms of this theory... but I didn't understand them... nor do I remember them... the magnitude argument is a bit confusing already.
I like this because it agrees that not everyone is born 100% gay or 100% straight... and that there is a little wiggle room for friskiness... and it makes a lot more sense than the arbitrary 1 in 10 theory.
The birth graph can be attributed to the amount of androgens the female bombards male fetuses (sorry girls, I don't know much about female homosexuality).
Journal of Theoretical Biology
The more androgens a female bombards... the more likely he'll be found on the 0-5 range. Read the Action of anti-male antibodies in the fetal brain chapter in the link above if you are interested.